73
Session One: Session One: Tax Systems and Taxpayer Behavior 2009 IRS Research Conference 2009 IRS Research Conference

Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Embed Size (px)

Citation preview

Page 1: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Session One:Session One:Tax Systems and Taxpayer y p yBehavior

2009 IRS Research Conference2009 IRS Research Conference

Page 2: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

M i th I t f T S tMeasuring the Impact of Tax Systems on Economic Behavior

Using New Cross-Country Data

Leslie Robinson, Tuck School of Business at DartmouthJoel Slemrod, University of Michigan

2009 IRS Research Conference

Page 3: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

MotivationMotivation

• Tax systems are multi-dimensional; e.g., rates and bases as stated may or may not be welland bases as stated may, or may not be, well administered and enforced.

E i ti t t di f th i t f• Existing cross-country studies of the impact of tax systems on behavior use measures of tax rates, and (sometimes) recognize the importance of administration and enforcementimportance of administration and enforcement using rough proxies.

• Leaving out important aspects of tax systems• Leaving out important aspects of tax systems may bias estimated partial effects of tax rates, and/or miss entirely the effects of other tax system features.y

Page 4: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Overview

• We construct 38 measures of tax system aspects y pfor 44 countries based on OECD (2006) and in this paper we examine 10 measures in 1 important context.context.

• We analyze the relation between new and existing proxies for tax administration and enforcement.p

• We re-examine the tax system as a determinant of the size of the informal economy (in the process y ( pwe search for exogenous variation in our measures).

Page 5: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Existing MeasuresEconomics Literature

• Tax rates are negatively associated with unofficial activity, g y y,while the tax burden is positively associated; WEF SURVEY (Johnson et al -AER 1998)

• Tax rates are not associated with unofficial activity after controlling for better-run administrations; GDP and LEGAL

( )INDEX (Friedman et al -JPE 2000)

• Tax rates are positively associated with unofficial activity, p y y,while tax enforcement is negatively associated; SEIGNIORAGE and LEGAL INDEX (Ihrig and Moe -JDE 2001 -AEJ 2004)

Page 6: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Existing MeasuresEconomics Literature

• Tax administration has a positive association with theTax administration has a positive association with the number of new business registrations; TAX PAYMENTS AND HOURS Djankovet al. (working paper, 2008)

• Tax rates have a negative association with self-l t h t f t i t demployment when tax enforcement is strong and no

association when tax enforcement is weak; CORRUPTION INDEX(Torrini -LE2005)INDEX (Torrini LE 2005)

Page 7: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Existing MeasuresFinance and Accounting Literature

• Tax enforcement has spillover effects on corporate p pgovernance; WEF SURVEY (Dyck and Zingales -JF 2004)

• Governance (but not tax enforcement) affects the relation between corporate tax rates and corporate p ptax revenues; WEF SURVEY (Desai et al. -JFE 2007)

T f t t i i t• Tax enforcement constrains earnings management; WEF /IMD SURVEY (Haw et al. -JAR 2004, Wysocki -JAR 2004))

Page 8: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

New Measures (OECD, 2006)Table 1

1 U fWi hh ldi 6 U f R i1. Use of Withholding– for 6 payment types

2 TaxSystem(forindividuals)

6. Use of Reporting– for 7 payment types

7. Self-Assessment2. Tax System (for individuals)– 4 classifications

3. Collect Power

– who computes tax liability?

8. Use of Matching– for 6 payment types

– out of 15 types of powers

4. Maximum Penaltyforfailuretocorrectlyreporttax

p y yp

9. Bank Access– strength of back secrecy laws– for failure to correctly report tax

liability

5. Administrator Coverage

strength of back secrecy laws and revenue agents’ ability to overcome them

10. Verification Power– revenue agent per 1,000 workers – search and seizure powers

Page 9: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

CorrelationsTable 2 (top)

r pita

om

e

Per

Cap

Inco

1 Per Capita Income 1 0000Per Capita Income 1.0000

Tax Hours -0.5371Tax Payments -0.4371

Corruption 0.8154Bribery 0.7543

Law and Order 0.7399Seigniorage to GDP -0.5827

Seigniorage -0.7281T C li 0 5944Tax Compliance 0.5944

Tax Burden 0.2682

Page 10: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

CorrelationsTable 2 (bottom)

r Cap

ita

com

e

x H

ours

x Pa

ymen

ts

rrup

tion

iber

y

w a

nd

der

gnio

rage

to

DP gn

iora

ge

x mpl

iane

x Bu

rden

Per

Inc

Tax

Tax

Cor

Bri

Law

Ord

Seig

GD

Seig

Tax

Com

Tax

1 2 3 4 5 6 7 8 9 10 Per Capita Income 1.0000 (0.5371) (0.4371) 0.8154 0.7543 0.7399 (0.5827) (0.7281) 0.5944 0.2682

Use of Withholding (0.4399) 0.4317 0.2278 (0.5533) (0.4796) (0.3090) 0.3258 0.5713 (0.4351) (0.0654) Administrator Coverage 0.3078 (0.3174) (0.1245) 0.2642 0.2505 0.4718 (0.3198) (0.6254) 0.0358 0.0886

Tax to GDP 0.4902 -0.3114 -0.3161 0.4268 0.4112 0.4519 -0.3634 -0.5586 -0.0075 -0.2133 Self Assessment -0.2851 0.2833 0.0205 -0.4015 -0.2907 -0.3010 0.3845 0.3532 -0.0622 -0.1618

System1 -0.3466 0.3305 0.0255 -0.3965 -0.2651 -0.2699 0.1374 0.2001 -0.1856 0.0140 System2 0.0198 -0.0062 0.1208 -0.0415 -0.1388 0.0150 0.1137 0.1240 -0.0641 -0.1727 System3 0.3365 -0.2653 -0.1752 0.4364 0.4111 0.3414 -0.2188 -0.3677 0.0996 0.0177 System4 0.1869 -0.2626 -0.0110 0.2603 0.2025 0.0426 -0.1469 -0.0840 0.3292 0.2339

Use of Reporting -0.2928 0.4461 0.0272 -0.3449 -0.3103 -0.1453 0.2246 0.6906 -0.0906 -0.0350 Collect Power -0.3733 0.1880 -0.0016 -0.3161 -0.3319 -0.2930 0.1237 0.2143 -0.4962 -0.3489

Max Penalty -0.3059 0.3612 0.1380 -0.1812 -0.2114 -0.2130 0.2642 0.3331 -0.0339 0.2123

Page 11: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

CorrelationsCorrelations

• Countries with strong legal systems /less corruption:– Less self-assessment of tax liabilities– Less use of withholding and information reporting

L t f ll ti f t d bt– Less power to enforce collection of tax debts– More tax administrators per worker

• Why should we characterize these countries as having strong tax administration and enforcement or low tax burdens?

Page 12: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

MAXIMUM PENALTY RATE ADMINISTRATOR COVERAGE PER CAPITA INCOME

LOWEST QUARTILE (1)

(2) Australia

(1)Chile

(2)Argentina

(1)Argentina

(2)Cyprus

Cyprus Finland Iceland Japan

Luxembourg

CanadaFrance

Hungary Norway

USA

ChinaJapan

Mexico Singapore

South Africa

Austria Cyprus

Italy Lithuania

New Zealand

ChileChina Latvia

Mexico Russia

HungaryItaly

Lithuania New Zealand

SingaporeLuxembourg Russia

South Korea Sweden

USA

South AfricaSouth Korea

USA

New Zealand Spain

Turkey

RussiaSouth Africa

Turkey

SingaporeSouth Korea

Spain

(3) HIGHEST (3) (4) (3) (4)Argentina

Latvia Lithuania Mexico

Netherlands

QUARTILE (4) Austria Belgium

Chile China

Australia Canada Finland France

Hungary

Belgium Denmark Iceland Ireland Latvia

Australia Belgium Finland France Japan

Austria Canada

Denmark Iceland IrelandNetherlands

New Zealand Spain

Turkey UK

ChinaDenmark Ireland

Italy Singapore

HungaryRussia

Sweden UK

Latvia Luxembourg Netherlands

Norway

JapanNetherlands

Sweden UK

IrelandLuxembourg

Norway USA

g pSouth Africa

Page 13: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

WITHHOLDING REPORTING PER CAPITA INCOME

LOWEST QUARTILE (1)

Australia Canada

k

(2)Austria Cyprus Finland

(1)Argentina

Austria Belgium

b

(2)Australia Canada Estonia

(1)Argentina

Chile China

(2)Cyprus

Czech RepublicGreece

Denmark France

Luxembourg Netherlands

Norway

GermanyIceland Malta

New Zealand Sweden

LuxembourgNetherlands

Slovak RepublicSwitzerland

France Germany Ireland Malta

New Zealand

EstoniaMexico Poland

Slovak RepublicSouth Africa

HungaryMalta

New Zealand Portugal SloveniaNorway

Singapore South Africa

USA

SwedenSwitzerland

New Zealand Norway

South Africa Sweden

South AfricaTurkey

SloveniaSouth Korea

Spain

(3) HIGHEST (3) (4) (3) (4)Argentina Belgium

Chile Czech Republic

Estonia

QUARTILE (4)China

Hungary Ireland Japan

DenmarkFinland Greece

Hungary Iceland

Chile China Cyprus

Czech RepublicJapan

AustraliaBelgium Finland France

Germany

AustriaCanada

Denmark Iceland IrelandEstonia

Greece Poland

Portugal Slovak Republic

JapanMexico

South Korea Spain

Turkey

IcelandMexico Poland

Singapore Slovenia

Japan Portugal

Spain Turkey

USA

GermanyJapan

Netherlands Singapore Sweden

IrelandLuxembourg

Norway Switzerland

USASlovenia

UK South Korea

UK UK

Page 14: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

WITHHOLDING REPORTING CORRUPTION

LOWEST QUARTILE (1)

Australia Canada

D k

(2) Austria Cyprus Finland

G

(1)Argentina

Austria Belgium

L b

(2)Australia Canada Estonia F

HIGH (1)Argentina

China Czech Republic

G

(2)Cyprus Estonia

Hungary JDenmark

France Luxembourg Netherlands

Norway

GermanyIceland Malta

New Zealand Sweden

LuxembourgNetherlands

Slovak RepublicSwitzerland

France Germany Ireland Malta

New Zealand

GreeceMexico Poland

Slovak RepublicSouth Africa

JapanMalta

Portugal Slovenia

South Koreao aySingapore

South Africa USA

S edeSwitzerland

e ea a dNorway

South Africa Sweden

Sou caTurkey

Sou o eaSpain

UK

(3) HIGHEST (3) (4) (3) LOW (4)Argentina Belgium

Chile Czech Republic

Estonia

QUARTILE (4)China

Hungary Ireland Japan

DenmarkFinland Greece

Hungary Iceland

Chile China Cyprus

Czech Republic Japan

AustriaBelgium Canada

Chile France

AustraliaDenmark Finland Iceland

New ZealandEstonia Greece Poland

Portugal Slovak Republic

JapanMexico

South Korea Spain

Turkey

IcelandMexico Poland

Singapore Slovenia

Japan Portugal

Spain Turkey

USA

FranceGermany Ireland

Luxembourg Netherlands

New ZealandNorway

Singapore Sweden

Switzerland Slovenia

UK South Korea

UK USA

Page 15: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

SELF-ASSESSMENT TAX SYSTEM

(INDIVIDUAL TAX)CORRUPTION

(INDIVIDUAL TAX)DO NOT USE

Austria Belgium Denmark Finland

SYSTEM 1Argentina

Austria Brazil Chile

SYSTEM 2 Australia Belgium Canada Greece

HIGH (1)Argentina

China Czech Republic

Greece

(2)Cyprus Estonia

Hungary Japan

Germany Greece Iceland

Luxembourg Netherlands

China Cyprus Czech

Republic Germany

Hungary India Malta Poland

Portugal

Mexico Poland

Slovak Republic South Africa

Turkey

pMalta

Portugal Slovenia

South Korea Spain

NorwaySingapore Slovenia

South Africa Sweden

IrelandItaly

Japan South Korea

Latvia Lithuania

Spain USA

UK

USE LithuaniaLuxembourg

Mexico Netherlands New Zealand

Russia

USEArgentina Australia Canada

Chile China

MaltaMexico

New Zealand Poland

Portugal

SYSTEM 3 Denmark Estonia Finland Iceland

(3)Austria Belgium Canada

Chile

LOW (4)Australia Denmark Finland Iceland

Slovak Republic Slovenia

South Africa Turkey

UK

China Cyprus

Czech Republic Estonia France

Hungary

PortugalSlovak Republic

Slovenia South Korea

Spain Switzerland

Iceland Norway Sweden

ChileFrance

Germany Ireland

Luxembourg Netherlands

IcelandNew Zealand

Norway Singapore Sweden

SwitzerlandSYSTEM 4

FranceUKIreland Japan

Turkey UK

USA

USA France

Singapore Switzerland

Page 16: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Informal Economy – Prior AnalysisInformal Economy Prior Analysis

Johnson et al (AER 1998) Freidman et al (JPE 2000)Johnson et al (AER 1998)

• Johnson et al (1998) regressed the level of IE against a measure of tax rates and an executive

Freidman et al (JPE 2000)

• Freidman et al (2000) regressed the level of IE against a measure

f “l d d ”of tax rates and an executive assessment of the tax burden, interpreted as “the way the tax system is administered.”

of tax rates, a “law and order” index and (sometimes) GDP per capita.

• Finding: High IEs are associated with low tax rates, and high tax burdens. The two RHS variables

• Finding: Tax rates have a negative affect on IE, but have no effect once one controls for “law and

d ” GDP ( hi h hare, however, not entered into the regression at the same time.

order” or GDP (which have a negative effect on IE).

Page 17: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

IV RegressionsTable 4

1 7 8Independent Variables

2 3 4 5 6FJKZ RS

0.3573 *** 0.4531 *** 0.4772 *** 0.5046 *** 0.4185 *** 0.5690 *** 0.5717 *** 0.6218 ***(0.0875) (0.0783) (0.0636) (0.0654) (0.0893) (0.0740) (0.0653) (0.0666)-0.3654 * 0.0789 0.3002 0.3481 -0.4786 ** -0.0866 -0.1690 -0.0871

(0.2012) (0.2170) (0.2149) (0.2137) (0.2193) (0.2876) (0.2607) (0.2732)-0.0932 *** -0.0377 -0.1027 *** -0.0547 *

Variables FJKZ RS

Intercept

Tax Rate

Log GDP Per (0.0243) (0.0232) (0.0344) (0.0298)

-0.0492 *** -0.0410 *** -0.0349 *** -0.0247 **(0.0097) (0.0110) (0.0118) (0.0118)

First stage F 4.13 3.71 5.28 4.58 8.75 8.42 10.99 8.29Adj R-sq 0.0690 0.3732 0.5391 0.5564 0.0881 0.3912 0.4467 0.4808

N 32 32 32 32 40 40 40 40

Law and Order

gCapita

• Columns 1-4– IVs are legal origin,

• Columns 5-8– IVs are conflict and

N 32 32 32 32 40 40 40 40

religion, latitude and ethnic fractionalization from La Porta et al (1999)

democracy from Besley and Persson (2007)

(1999)

Page 18: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Informal Economy – New AnalysisInformal Economy New Analysis

• Tax rates have no significant effect on IE, while h fnew tax system measures have significant

positive and negative effects.

• In some specifications, per capita income and the legal environment have no significant effect.g g

• Suggests that the relationship between income d IE l f h iand IE may result from tax system choices

shaped by other institutional factors that are correlated with income.

Page 19: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

IV RegressionsT bl 5Table 5

Intercept 0.7708 *** 0.6972 *** 0.0574 0.3286 0.4761 ***(0.0918) (0.0622) (0.3332) (0.1981) (0.1256)

Tax Rate -0.3494 0.0287 -0.4263 -0.3712 -0.2507(0 21 3) (0 216 ) (0 4220) (0 32 3) (0 2 90)(0.2173) (0.2167) (0.4220) (0.3253) (0.2590)

Log GDP Per Capita -0.0860 ** -0.0986 *** 0.0074 -0.0085 -0.0294(0.0330) (0.0317) (0.0666) (0.0212) (0.0317)

Law and Order -0.0093 -0.0355 *** -0.0133 -0.0103 -0.0248 **(0.0108) (0.0123) (0.0137) (0.0444) (0.0109)

M P lt 0 0524 **Max Penalty -0.0524 **(0.0204)

Administrator Coverage 0.0710 **(0.0267)

Collection Power 0.0396 *(0 0228)(0.0228)

Use of Withholding 0.0389 *(0.0210)

System1 0.1580 *(0.0874)

System2 0 1034System2 0.1034(0.0829)

System3 0.1928 **(0.0850)

1.872.291st stage F - System Measure 2.29

3.97 4.96 3.20 2.44 13.971st stage F - Tax Rate 6.59 7.81 5.60 5.56 4.32

Adjusted R-sq 0.5644 0.6210 0.4157 0.3753 0.5171 Sample 30 39 37 32 40

1st stage F System Measure

Page 20: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

ConclusionConclusion

• Tax system aspects – other than rates – matterTax system aspects other than rates matter.

i h k ’ i f h i fl f• Ignoring them skews one’s view of the influence of tax rates on behavior.

• Unexplored variables await the attention of future presearch.

Page 21: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Session One:Session One:Tax Systems and Taxpayer y p yBehavior

2009 IRS Research Conference2009 IRS Research Conference

Page 22: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

A PANEL ANALYSIS OF BEHAVIOUR CHANGE IN CANADIAN INDIVIDUAL INCOME TAX COMPLIANCEINDIVIDUAL INCOME TAX COMPLIANCE

Presented by Attah Boame July 8, 2009

2009 IRS Research Conference, Washington, D.C., USA

Page 23: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

OUTLINEBackground and Objectives

Methodology

Study Results

fAreas for Further Research

Questions or Comments?Questions or Comments?

Page 24: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Background:Personal income taxes are a major source of income for the federal, provincial and territorial governmentsgovernments.Personal income taxes generated an average of $81.6 billion quarterly from 2000 to 2004 for q ythe federal government.

St d bj tiStudy objectives:To study the trends in Canadian individual yincome tax compliance from 1996 to 2002.To analyze the factors that influence individual income tax compliance for the study periodincome tax compliance for the study period.

Page 25: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Tax Compliance Definedp

The Compliance Measurement Framework (CMF) identifies the following main compliance requirements for individuals:

Filing required tax forms on time;

Reporting complete and accurate tax information; and

Paying any amounts due in a timely manner (without enforcement action).

Page 26: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Operational DefinitionT C liTax Compliance

Filing Compliance Rate - The number of g Co p a ce ate e u be otaxpayers filing on-time (i.e., with no late filing penalty) as a percentage of the panel population.Reporting Compliance Rate - The number of taxpayers reporting accurately (i.e., with a tax payable difference of less than or equal to $50payable difference of less than or equal to $50 between assessment and what is reported) as a percentage of the panel population.P C li R Th b fPayment Compliance Rate - The number of taxpayers without arrears interest charges or instalment interest charges as a percentage of the panel populationthe panel population

Page 27: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

DataSources: Initial Assessment and Reassessment of individual taxpayers tax returns (T1)of individual taxpayers tax returns (T1)

Study period: From 1996 to 2002

Variables: As defined in the datasets

Type: Panel (Longitudinal) datayp ( g )• To achieve a balanced panel, only individual

taxpayers who filed their tax returns for all seven years are included in the analysis.y y

Observations:18,300,485 for each tax year

Page 28: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

MethodologyMethodologyDescriptive Statistics

F di t ib ti d C t b l ti• Frequency distributions and Cross-tabulations

Multivariate Analysisy• Logistic regression (identifies the likelihood of the

individual taxpayers filing their taxes on-time, reporting their taxes accurately, and paying their p g y, p y gtaxes owing on-time, without enforcement action by the Agency).

Page 29: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

CaveatsNon-filers are excluded from the analysis.Taxpayers with no tax payable are included in order to preserve the balanced structure of theorder to preserve the balanced structure of the data.The data are applicable to most T4 recipients (that is employees)(that is, employees).The number of observations for each tax year represent about 80% of the filing population for each tax yeareach tax year.Taxpayers over-reporting their tax payable are considered reporting compliant. Th l i ti i d lThe logistic regression models assume no interaction between variables (weak dependencies identified among independent variables)variables).

Page 30: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Study ResultsCompliance trends:

• Reporting compliance rate exceeded both filing and p g p gpayment compliance rates for the study period. Reporting compliance rate decreased from 96% in 1996 to 95% in 2002.

• Filing compliance rate exceeded payment compliance rate for all the years of the study period. Filing compliance rate increased slightly from roughly 92% in 1996 to 93% in 2002in 1996 to 93% in 2002.

• Payment compliance rate was lower than both filing and reporting compliance over the study period.

Page 31: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Filing, Reporting and Payment Compliance, 1996-2002

98

94

96

%)

90

92

Rat

e (%

86

88

1996 1997 1998 1999 2000 2001 20021996 1997 1998 1999 2000 2001 2002

Filing Reporting Payment

Page 32: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Study ResultsyDemographic trends:

• Females are more tax compliant (filing, reporting, and payment) than males.

• Middle age taxpayers are less tax compliant (filing, reporting, and payment) compared to the young and old taxpayers.

• Widowed taxpayers are more tax compliant (filing and reporting) than those in the other marital status

t icategories.

Page 33: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Canada, Filing Compliance by Gender (%), 1996-2002

93

92.5%)

92Rat

e (

91.51996 1997 1998 1999 2000 2001 20021996 1997 1998 1999 2000 2001 2002

Canada Male Female

Page 34: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Canada, Reporting Compliance by Gender (%), 1996-2002

97.5

96.5

%)

94.5

95.5

Rat

e (

93.51996 1997 1998 1999 2000 2001 2002

Canada Male Female

Page 35: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Canada, Payment Compliance by Gender (%), 1996-2002, y p y ( ),

90.5

92.5

%)

86.5

88.5

Rat

e (

84.5

86.5

1996 1997 1998 1999 2000 2001 20021996 1997 1998 1999 2000 2001 2002

Canada Male Female

Page 36: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Mean Compliance By Age Group (%),1996-2002

100.00

92.00

96.00

%)

84 00

88.00Rat

e (

80.00

84.00

0 14 15 24 25 34 35 44 45 54 55 64 65+0 - 14 15 - 24 25 - 34 35 - 44 45 - 54 55 - 64 65+

Filing Reporting Payment

Page 37: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Mean Compliance by Marital Status (%),1996-2002

98.00

90.00

94.00

(%)

82.00

86.00Rat

e

78.00Married Common- Widow ed Divorced Separated Single

Law

Filing Reporting Payment

Page 38: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Study Results yFiling Methods:• Taxpayers using electronic methods

(Efile, Telefile, and Netfile) have a hi h li t (fili tihigher compliance rate (filing, reporting, and payment) relative to taxpayers who use paper-filing (hard copy)use paper filing (hard copy).

Page 39: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Mean Compliance by Filing Method (%), 1996-2002

90 00100.00

p y g (%),

70.0080.0090.00

(%)

40.0050.0060.00

Rat

e

20.0030.00

Paper Efile Telefile EDI Internet HomeFiliFiling

Filing Reporting Payment

Page 40: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Study ResultsyMarginal Tax Rates:

Th li d li i h h• The payment compliance rate declines with the marginal tax rates. That is, the lower marginal tax bracket (16%) has a higher payment compliance rate relative to the other marginal tax brackets (22% 26%relative to the other marginal tax brackets (22%, 26% and 29%).

• The lower marginal tax bracket (16%) has the highest g ( ) greporting compliance rate compared to the other marginal tax brackets (22%, 26% and 29%).

Page 41: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Mean Compliance by Marginal Tax Rates (%), 1996-20021996-2002

85 0090.0095.00

%)

70 0075.0080.0085.00

Rat

e (%

60.0065.0070.00

MTR (16%) MTR (22%) MTR (26%) MTR (29%)MTR (16%) MTR (22%) MTR (26%) MTR (29%)

Filing Reporting Payment

Page 42: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Logistic Regression Resultsg g

Odds ratio analysis:Odds ratio analysis:• The Odds ratio indicates how much

more likely a certain event occurs in one ygroup relative to its occurrence in another group, all other things being the same For example how much moresame. For example, how much more likely are females (reference category) filing compliant compared to males?g p p

Page 43: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

P t Fili C li O tParameters Filing Compliance OutcomeDemographic Factors Less Likely

to File LateMore Likely to File Late

Gender Male vs. Female 11%

Age Group Middle vs. Young 8%Old vs. Young 106%

Marital Status Widowed vs.Married/CL 26%Divorced vs.Married/CL 39%Separated vsSeparated vs.Married/CL 45%Single vs.Married/CL 27%Married/CL 27%

Page 44: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Parameters Filing Compliance Outcome

CRA Program Factors Less Likely to File Late

More Likely to File Late

Filing Method EFILE vs. Paper 385%g p

TELEFILE vs.Paper 603%

NETFILE vs.Paper 204%

Marginal Tax 22% bracket vsMarginal TaxRates (%)

22% bracket vs.16% bracket 5%

26% bracket vs.16% bracket 16%

29% bracket vs.16% bracket 85%16% bracket 85%

Page 45: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Parameters Filing Compliance OutcomeIncome Factors Less Likely

to File LateMore Likely to File Late

Pension Income Pension vs. NoPension 109%

RRSP Income Income vs. NoIncome 18%

Tax-ExemptIncome

Exempt vs. NoExempt 20%

Main Source of Investment/RentMain Source ofIncome

Investment/Rentvs. Wages 29%CapitalGains/Loss vsGains/Loss vs.Wages 32%Self-Employedvs Wages 2%vs. Wages 2%

Page 46: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

P t R ti C li O tParameters Reporting Compliance OutcomeDemographic Factors Less Likely to

UnderreportMore Likely to Underreport

Gender Male vs. Female 33%Age Group Middle vs. Young 13%

Old vs Young 8%Old vs. Young 8%Marital Status Widowed vs.

Married/CL 10%Divorced vs.Married/CL 25%Separated vs.Married/CL 29%Single vs.Married/CL 4%

Page 47: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Parameters Reporting Compliance Outcome

CRA Program Factors Less Likely to More Likely to Underreport Underreport

Filing Method EFILE vs.Paper 152%NETFILE vs.Paper 133%

Marginal Tax 22% bracket vsMarginal TaxRates (%)

22% bracket vs.16% bracket 13%26% bracket vs.16% bracket 11%16% bracket 11%29% bracket vs.16% bracket 4%

Page 48: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Parameters Reporting Compliance Outcome

Income Factors Less Likely to Underreport

More Likely to Underreport

Pension Pension vs NoPensionIncome

Pension vs. NoPension 22%

RRSPIncome

Income vs. NoIncome 58%Income Income 58%

Tax-ExemptIncome

Exempt vs. NoExempt 178%

M i S I t t/R tMain Sourceof Income

Investment/Rentvs. Wages 3%CapitalGains/Loss vs.Wages 40%Self-Employedvs. Wages 17%

Page 49: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Parameters Payment Compliance Outcome

Demographic Factors Less Likely to Pay Late

More Likely to Pay Latey y

Gender Male vs. Female 30%Age Group Middle vs. Young 6%

Old Y 20%Old vs. Young 20%Marital Status Widowed vs.

Married/CL 19%Divorced vs.Married/CL 30%Separated vs.Separated vs.Married/CL 33%Single vs.Married/CL 2%Married/CL 2%

Page 50: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Parameters Payment Compliance OutcomeParameters Payment Compliance OutcomeCRA Program Factors Less Likely to

Pay LateMore Likely to Pay Late

Filing Method EFILE vs. Paper 170%TELEFILE vs.Paper 1,310%NETFILE vs.Paper 333%

Marginal Tax 22% bracket vsMarginal TaxRates (%)

22% bracket vs.16% bracket 35%26% bracket vs.16% bracket 51%16% bracket 51%29% bracket vs.16% bracket 60%

Page 51: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Parameters Payment Compliance OutcomeOutcome

Income Factors Less Likely to Pay Late

More Likely to Pay Late

Pension Income Pension vs. NoPension 27%

RRSP Income Income vs. NoIncome 60%

Tax-ExemptIncome

Exempt vs. NoExempt 138%

Main Source ofIncome

Investment/Rentvs. Wages 28%Capital Gains/LossCapital Gains/Lossvs. Wages 41%Self-Employed vs.Wages 64%Wages 64%

Page 52: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Further ResearchFurther ResearchA profile and compliance trends for:

• Taxpayers who use professional tax preparers to file their tax return;preparers to file their tax return;

• Taxpayers who use the Community Volunteer Income Tax Program to file gtheir tax return;

• Taxable filers;• Non-taxable filers; and • Refund returns.

Page 53: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

QUESTIONS or COMMENTS?

THANK YOU

MERCI

Page 54: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Session One:Session One:Tax Systems and Taxpayer y p yBehavior

2009 IRS Research Conference2009 IRS Research Conference

Page 55: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Taxpayer response to the recent 'flat tax' UK capital gains tax (CGT) reform and theUK capital gains tax (CGT) reform and the implication for corporate tax simplification

P t J lfPeter JelfsMazars LLP

Page 56: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Introduction• UK corporate tax law is complex – the longest

in the world• Taxpayers and government both claim to want

it simplified• Every year the legislation increases in length

and complexity – why?• Taxpayer response to the recent CGT reforms

may help to explain

IRS Research Conference 2009

Page 57: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

UK corporation tax• Companies taxed separately from their

shareholders• Controlled by statute, case law and practices

of HM Revenue & Customs (HMRC)• Legislation constantly updated• Responsibility usually delegated by

government to HMRC

IRS Research Conference 2009

Page 58: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Reasons for complexity• Complexity can arise through length of

legislation or language usedA b f ifi f t i th UK• A number of specific factors in the UK:

Conflict bet een polic goals• Conflict between policy goals- often equity and simplicity

• Fiscal incentives- create distinctions- create distinctions

IRS Research Conference 2009

Page 59: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Reasons for complexity (2)• Anti-avoidance legislation

- loopholes created and exploited

• Tax Law Rewrite Projectj- designed to simplify language

IRS Research Conference 2009

Page 60: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Reasons for complexity (3)Other reasons:• Relieving capital expenditureg• Tax neutrality• Earlier shortcomingsg• True reflection• Transfer pricingTransfer pricing

IRS Research Conference 2009

Page 61: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Effects of complexity• On first principles length of legislation can measure complexity• UK tax law increased from 5952 pages in 2001 to 10134 in 2008• Cost/benefit analysis for obtaining tax advice• Compliance falls through ignorance rather than evasion• Several surveys conclude taxpayers find

t t lsystem too complex

IRS Research Conference 2009

Page 62: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

‘Flat tax’• Two kinds – Hall Rabushka (HR) and Eastern European (EE)• HR a theoretical consumption based tax – not well known in UK• Flat tax in UK means EE• EE flat tax retains income base but has single rate of tax

IRS Research Conference 2009

Page 63: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Flat tax and simplification• Debate in UK as to whether flat tax would simplify legislation• Possible removal of capital/income planning and fewer exemptions• Effect on length of legislation predicted to be small

IRS Research Conference 2009

Page 64: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

CGT reforms• 9 October 2007 – flat tax rate of 18% on capital gains for individuals• Substantial simplification of CGT legislation• Removal of complex calculations of taper relief reducing gain depending on time asset held and whether used in business• Taper relief described as ‘a mess’ – range of CGT rates from 5-40%

Di t t d i t t d i i• Distorted investment decisions

IRS Research Conference 2009

Page 65: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

CGT reforms(2)• UK government does not like ‘flat tax’• Critical report in 2005 mainly on equity grounds y y g– reduction in marginal rate for highest earners• Report considered important the fact that no flat tax introduced in a Western economy• Government introduced this reform with flat tax simplification arguments – ironic!• How will taxpayer respond?

IRS Research Conference 2009

Page 66: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

CGT reforms(3)• Bitterly denounced by small business representatives!• Reason – flat rate was higher than the current 10%• Tax practitioners – pragmatic approach by schemes to avoid new rates• Simplicity a principle easily sacrificed

IRS Research Conference 2009

Page 67: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Government response• Quickly gave in• Introduced ‘entrepreneur’s relief’ to maintain 10% rate on first £1 million of capital gain –remainder at 18%• Simplification implications – disastrous• No simpler than taper relief and less generous

IRS Research Conference 2009

Page 68: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Conclusions• Classic example of taxpayer response to a radical system change• Rate of tax is the key factor• Simplification will remain elusive

IRS Research Conference 2009

Page 69: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference
Page 70: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Session One:Session One:Tax Systems and Taxpayer y p yBehavior

2009 IRS Research Conference2009 IRS Research Conference

Page 71: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Tax Systems and Taxpayer y p yBehavior

Discussant: Pamela Olson

Sk dd A Sl t M h & Fl LLPSkadden, Arps, Slate, Meagher & Flom LLP

2009 IRS Research Conference2009 IRS Research Conference

Page 72: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

Session One:Session One:Tax Systems and Taxpayer y p yBehavior

2009 IRS Research Conference2009 IRS Research Conference

Page 73: Session One:Session One - Internal Revenue Service | An ... · Session One:Session One: Tax Syypystems and Taxpayer Behavior 2009 IRS Research Conference2009 IRS Research Conference

During the break, please visit the SOI Booth in the upperthe SOI Booth in the upper lobby.

2009 IRS Research Conference2009 IRS Research Conference